Forum Discussion

moog67_108621's avatar
moog67_108621
Icon for Nimbostratus rankNimbostratus
May 14, 2012

Big-IP reboot unexpectedly

Hi everyone,

 

 

Our BIG-IP 1600 system rebooted unexpectedly after a "bigstart restart" command was launched by the HA monitor processes.

 

 

The following log can be seen in ltm.log

 

 

May 11 04:07:20 local/xxxxxxx notice overdog[3163]: 01140029:5: HA daemon_heartbeat mcpd fails action is restart all.

 

May 11 04:07:20 local/xxxxxxx crit overdog[3163]: 01140106:2: Overdog daemon calling bigstart restart.

 

May 11 04:07:20 local/xxxxxxx notice sod[3297]: 010c0044:5: Command: go standby overdog.

 

May 11 04:07:43 local/xxxxxxx notice logger: /usr/bin/overdog ==> /usr/bin/bigstart restart

 

May 11 04:07:59 local/xxxxxxx notice overdog[3163]: 01140030:5: HA daemon_heartbeat mcpd is now responding.

 

May 11 04:09:41 boot_marker: ---===[ HD1.1 - BIG-IP 10.2.0 Build 1707.0 ]===---

 

May 11 04:09:43 local/xxxxxxx info mprov: Invoked as: /usr/bin/mprov.pl (pid=1812) --boot --quiet

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/gtm

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/tmm

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/wom

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/psm

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/ui

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/lc

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/woml

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/apm

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/wam

 

May 11 04:09:43 local/xxxxxxx info mprov: /bin/mkdir -p /dev/mprov/asm

 

May 11 04:09:43 local/xxxxxxx info mprov: No mountpoints existed, likely first boot after upgrade.

 

 

We are running 10.2.0 version.

 

 

Any ideas why this happened?,

 

 

Thanks in advance,

 

 

  • nathe's avatar
    nathe
    Icon for Cirrocumulus rankCirrocumulus
    moog67

     

     

    Has this happened just once? I see it happened around 4am - this is when the ltm syslog service does a reload of the configuration daily (you can double check this is var/log/messages). It looks like the offending daemaon is the mcpd one - and this manages the config on the box...amongst other things.

     

     

    Does it crash if you update the config or do a config sync?

     

     

    I'm not sure exactly what might cause this but there is a Known Issue with the mcpd daemon and ltm v10.2.0 - see http://support.f5.com/kb/en-us/solutions/public/12000/100/sol12176.html so I would think about applying a hotfix to see if that sorts the issue.

     

     

    Hope this helps,

     

    N

     

     

  • Nathan hi,

     

    Thanks for the quick reply.

     

    We have just seen it once so far, and when doing a config sync both boxes (Active/Standby) stay up and running.

     

     

    We'll keep an eye on the mcpd memory leaking to see if it is the same case.

     

     

    Regards,

     

    moog67